Independent predictors of survival in patients with incurable malignant gastric outlet obstruction: a multicenter prospective observational study

Abstract

Objective: Gastric outlet obstruction (GOO) is one of the late complications of a variety of malignancies. Palliation of symptoms of obstruction rather than cure is the primary aim of treatment in affected patients. Thus far prognostic information on life expectancy is lacking in these patients although it can be of importance when deciding upon their optimal treatment. The purpose of this study was to investigate whether baseline data in patients with incurable GOO can independently predict survival.

Patients and methods: In total, 105 consecutive patients with symptomatic GOO treated with duodenal stent placement were enrolled in this multicenter prospective observational study. Patients were followed until death or till 1 November 2008. The Cox proportional hazard regression model was used for both univariate and multivariate analyses of survival.

Results: Baseline data of 101 patients were completed. At the time of analysis, 95% of patients had died; median overall survival was 82 days (75% alive at 36 days, 25% alive at 156 days). The final prediction model revealed the dichotomized WHO performance status (HR: 2.63, 95% CI: 1.68-4.12, p < 0.001), prescription of morphines stronger than tramadol (HR: 2.42, 95% CI: 1.38-4.25, p = 0.002) and pain score of the EORTC QLQ-C30 (HR: 1.01, 95% CI: 1.00-1.01, p = 0.035) as independent significant prognostic factors for short survival.

Conclusions: This study demonstrates clear predictors of poor outcome for patients presenting with symptomatic malignant GOO. The model may enhance the selection of optimal treatment for individual patients.